Shu cases studies

This repository contains notebooks for reproducing the two figures in the manuscript Shu: Visualization of high dimensional Biological Pathways.

Figure 1: maud case study

The kinetic model can be found at models/maud_ecoli_glycolysis_and_ppp. The output of Maud is provided at data/maud_output.

To reproduce the output generation (might take a day):

# install the bayesian kinetic modeling software
pip install maud-metabolic-models==0.4.0.2
install_cmdstan
# sample a posteriori distribution
maud sample models/maud_ecoli_glycolysis_and_ppp

The notebook notebooks/1_shu_maud_kcat_priors.ipynb performs the required data wrangling using pandas (without ggshu) to produce the input for shu (click on "Data" at https://biosustain.github.com/shu) at data/maud_output.metabolism.json. The network map is at maps/ecoli_map.json.

Figure 2: omics case study

The notebook notebooks/2_shu_omics.ipynb arranges the omics data, computes the flux sampling (using models/iCLAU786.xml and models/eciICLAU786.xml) and uses ggshu to produce the input for shu at data/omics.metabolism.json.

The map is a custom one for Clostridium autoethanogenum and is available at data/cauto_map.json.